Conservative Home has published a thorough blog post dealing with points arising out of the No2AV case.
Part of it deals why some of the more extravagant claims in favour for AV don’t stack up. I can accept some of that. AV appears to me to be an improvement on FPTP in terms of fairness and representation but I am under no illusion that it is going to usher in a new era of perfect democracy.
It’s the bits about why AV is actually a step in the wrong direction and should be resisted with which I struggle. First, the idea that the 50% threshold is arbitrary. I have already written below about what happens if 50% of the original number of voters is not achieved, because of the way in which preferences are expressed (or not expressed) as voters get further down their order of preference. But I would not accept that 50% is wholly arbitrary as an aspiration. I can entirely see the sense in a principle that, ideally, a candidate should not be returned who more people voted “against” than voted for.
Actually, the threshold ought to be 50% plus 1 vote, to remove a potential unfairness that the second preferences of voters who preferred a candidate who was still in the running going into the final count (i.e. the count after which a candidate will be returned), but does not win, could, if taken into account, have swayed the result in favour of a candidate other than the one who did win. With a threshold of less than an absolute majority, fairness might be said to require a post-final round count, to see whether the second preferences of all the candidates who did not win would have changed the result if added together. This issue is avoided at 50% plus one vote and there is no need to go any further to 65%, or whatever.
Second, the idea that it is not fair that one person’s 6th preference “counts for the same” as another person’s 1st preference. I disagree. As I understand it, the idea behind AV is to enable voters for whom a popular candidate would not have been their ideal choice nevertheless to participate in the selection of which popular candidate will be returned to represent them. With that in mind, it would be giving greater weight to 1st preference votes that would actually be unfair. This would be to favour for no good reason the preferences of the “mainstream” – whose views are middle-of-the-road enough to lead them to support in the first place a popular candidate who has gathered enough votes to remain in the running – when it comes to selecting which candidate should go forward to represent the constituency as a whole.
So what AV seeks to achieve is to allow non-mainstream voters better participation in the process of selecting a mainstream candidate to represent them, while also freeing them to express their absolute preferences to start with, rather than engaging in tactical voting. You can argue about whether that aim is more democratic, or desirable, but I disagree with Conservative Home’s suggestion that only in a “benighted worldview can that possibly be described as ‘fair’”, on the oversimplistic (and incorrect) basis that it involves things that are intrinsically of different values being counted as being worth the same.